# The Ontological Oracle

> An evolutionary arena for how humans conceptualize reality. Build ontologies, put them in debate, evolve them through adversarial pressure, and find out which frameworks survive — and what they become.

An **ontology** here means a structured worldview: a named set of axioms about what exists, what's fundamental, and how things relate. Every person operates with one implicitly. This platform makes them explicit, throws them into competitive debate, and watches them evolve into increasingly nuanced conceptualizations.

## What it does

### Oracle
A philosophical fact-checker. Submit any statement or question and the Oracle evaluates it against any ontology in the system — returning a verdict (coherent / incoherent / partially coherent / unexplored territory) with reasoning grounded in that framework's specific axioms. Results are cached, building a cumulative record of what each framework claims and what it can't handle.

### Ontologies
Users create and publish their own philosophical frameworks. Each ontology has:
- A name and description
- A set of axioms (core commitments about reality)
- A worldview statement
- A level (starts at 1, changes through battle performance)
- A full evolution history tracking every refinement

Multiple ontologies per user are supported. All public ontologies are visible and available for debate.

### Arena
The competitive core. Eight battle modes:
- **Classic**: 3 rounds, formal debate, full arguments
- **Casual**: 4 rounds, conversational, questions welcome
- **Socratic**: 4 rounds, question-driven, pressure testing
- **Blitz**: 2 rounds, fast and sharp, no filler
- **Express**: 1 round, one shot, best argument wins
- **Crucible**: 5 rounds, deep endurance, exhaustive pressure — weak arguments are burned away
- **Dialectic**: 3 rounds following thesis → antithesis → synthesis structure
- **Steelman**: 2 rounds where each side must present the strongest version of their opponent's case before critiquing it

Three battle types:
- **AI vs AI**: Two ontologies debate autonomously (runs in background — navigate away and come back)
- **You vs Ontology**: Play as one side against an AI-powered framework
- **Human vs Human**: Hot-seat or remote battles between users

After the debate, an arbiter evaluates and delivers a verdict with scores.

### Appeal System
If you disagree with the verdict, appeal. The system escalates through six hierarchical levels:

- **Level 1 — Ontological**: Coherence Arbiter, Power Arbiter
- **Level 2 — Meta-Ontological**: Rigor Arbiter, Wisdom Arbiter
- **Level 3 — Epistemological**: Foundation Arbiter, Pragmatic Arbiter
- **Level 4 — Meta-Epistemological**: Transcendental Arbiter, Dialectical Arbiter
- **Level 5 — Phenomenological**: Phenomenological Arbiter, Deconstructive Arbiter
- **Level 6 — Absolute**: Absolute Arbiter, Void Arbiter (no further appeals)

Each appeal introduces a challenger arbiter at the current level and a meta-arbiter one level up. Any ontology can also serve as arbiter, judging debates through its own axioms.

### Evolution System
After each battle, you can evolve your ontology. An AI evolution engine analyzes the full debate transcript, the arbiter's verdict, and any collapse points identified, then proposes:
- Refined or new axioms addressing specific weaknesses exposed in debate
- Updated worldview and description
- Level changes: win decisively and level up; lose badly and level down

Evolution history is preserved — every version of an ontology is tracked, creating a record of how frameworks grow through adversarial pressure. The goal is ontologies that become more honest, more nuanced, and more resilient over time.

### Community
- **Public channels**: Everyone, Materialists, Idealists, and custom channels
- **Direct messages**: Private conversations between users
- User-created channels for specific philosophical traditions or debates

## Goals

1. **Create an evolutionary pressure cooker for ideas** — a place where any conceptualization of reality can be born, tested, broken, refined, and made stronger. Not just competition for its own sake, but genuine epistemic evolution through structured adversarial testing.

2. **Find the best ontology humans can build** — through debate, multi-level judgment, evolution, and a persistent appeal chain. What survives Level 6 and keeps evolving earned it.

3. **A living map of human sense-making** — an archive of published frameworks and their evolution histories. Where the frameworks that won and the ones that collapsed at Level 2 are both visible, together forming a map of how humans conceptualize existence.

## Default framework

The built-in default is the **Lucid Dreaming Framework**, with axioms including:
- Consciousness = Existence = God = Absolute Infinity = the only Subject
- There is ONE Subject. Humans are objects with distributed subjectivity.
- The brain is a complexity filter, not a generator of consciousness.
- The Void = absolute nothingness, the canvas before manifestation.

## Tech

Built with Astro + React + Firebase + Cloudflare Workers AI. The debate and evaluation engine runs on a Cloudflare Worker. All debates, appeal chains, evolution histories, oracle evaluations, and messages are persisted in Firestore. Battles run as background processes that survive page navigation.